Afterimage

I use reddit a fair bit and am well aware how often they are completely off the mark or missing the bigger picture. However after trying to analyze this tweet for an hour or so I don't blame them at all.

The tweet is contradictory, inflammatory and extremely biased.

If the central point is that China's motive for open source is to scare off investment in frontier training then it is baffling that he specifically lays out China's reason for open source -

"I suspect the reason they are is 75% explained by strategic blindness/lack of AGI-pilledness (the CCP ... (read more)

I like the idea. I'm sure there is a large number of people, who would be interested in these ideas but hasn't been exposed to them. I would assume that pamphlets are used already but if they're not that's a massive advertising gap.

I get your point, I'm also frustrated that very few people seem to be aware that the likelihood of something insane happening in the next 10 years seems quite high. Especially for me I'm very unused to being the most alarmed person in the room so it's easy to feel annoying or crazy for trying to talk about these things. I suppose other people with a past history of passionate causes are used to these feelings and to them I seemed like the one with my head stuck in the sand.

If it makes you feel better, if you (we) are right, there will come a point where t... (read more)

Fair enough, I haven't seen that link before, interesting stuff. You've inspired me to try an extra high protein diet and see how I feel. Interestingly the USDA's new guideline is 1.2g-1.6g/kg and that's not even for strength training.

I think the only thing I'd change with the framing is that I got the impression 1.6g-2.2g/kg was required for strength training rather than what you need to maximize gains. Otherwise I retract my critic of the science, the evidence does seem to point towards 2g/kg or perhaps more as being optimal.

Great post, thanks for posting. Since you asked for suggestions I'll give a nitpik,

- "1.6-2.2g/kg in order to preserve or build muscle."

1.6g/kg seems to be where most people fail to see gains from additional consumption of protein, you can build muscle while eating 0.8g/kg it just potentially won't be as quick. So dont hold yourself back from working out just because you can't/won't change your diet to be high protein, you'll still get large gains, especially if you are untrained (both from the neural side and the muscle growth side).

To be clear I still... (read more)

Trying to use LLM's to back up your arguments in real time is insane. This seems self-evident to me, but I'll briefly list a few of the issues.

Hallucinations, bias from the prompt, bias from memory/custom instructions, bias from the specific model, having to read a long response just to see if the answer is a good one, LLMs just sometimes having poor reasoning skills, LLM's not focusing on the point you were trying to get across, lack of transparency with what you've pre-promoted the LLM with, I could go on.

As a layperson MV came across quite unconvincin... (read more)

I'm curious why you think that, unless it only applies to America. In my culture unions are seen quite positively.

I think that's an unfair deployment of xkcd mockery. 8% of a pandemic this year is not a tiny chance, which means a 50% drop is actually a big deal. The issue was interpreting the prediction market as an accurate percentage when it should just be an indication of approximate risk.

Wouldn't that imply that Opus 4.6 should have been below trend since it wasn't a step change in model size or pretraining (as far as I know).

Thanks for posting, I find these updates very interesting. How will you know when the Automated Coder milestone is reached? It seems like there weren't be a point where it's clear AI companies would prefer to lay off their workers rather than stop using AI. Is there also a chance that AI could 10x or 100x development speeds before it is possible for the software engineers to be laid off (if for example one of the things currently holding it back is unable to solved), would you consider that automated coding?

I'm curious are people disagreeing with the points about centralisation > decentralisation, or the points about the usa being the best/better than china. Assuming that USA or china is the country to develop TAI, (free feel to correct that assumption if you disagree about the likelihood), as a non-american/non-chinese person I'm not happy with either option but still think the USA has the edge in creating a better post-TAI world than China would. I wouldn't have thought this claim was so blantently false as to gather 23 disagree points.  

As a data point after reading your post and the apple comic I still don't understand how primer time travel works. If forced to write a fanfic I think I would butcher it too. I'm not sure if I should be holding an LLM to a higher or lower standard than myself. 

The examples are clearly favoring haiku for those who want a summary. 

-Chatgpt: 674 words to conclude there is no clean answer. 

Haiiku - 274 words and gives an unambiguous answer. 

I have read your link and I understand you seem to have much more knowledge about statistics than I am. Perhaps I'm making a simple mistake somewhere in my reasoning. My thoughts are:

 50 success horizon went from 1 to 4 hours. I interpreted your comment and post to posit that the increase is to some large degree due to training on cybersecurity to increase preformance in the 2-4hr range. However, looking at the charts it seems clear that the increase is not dominated by increased preformance in the 2-4hr range.  

My point isn't that there's enough... (read more)

If I'm intrepreting these charts correctly, there is a decent amount of progress in 15m-1hr bracket, a small amount of progress in the 1hr bracket, a decent amount of progress in the 2-4 hr bracket, and a large amount of progress in the 4-16hr bracket. It doesn't look like progress was dominated by the 1-4hr range.

Thought-provoking post, especially the section about the meaning of traditional food. I'm curious about the relationship between insulin resistance and obesity. If you can keep skinny can you avoid the metabolism FUBAR you're talking about even if your diet is still westernised? 

So, I have to admit I'm still confused. Is the icecream example fairly unrelated to the introduction and first chapter? They seem to be talking mostly about pure qualia, while the ice-cream example is talking about actions. 

I agree qualia is entirely disconnected from rationality, but I think anything beyond qualia such as actions or intending to take actions is rationality fair game so to speak. I don't see an issue in Bryce assessing the rationality of Ash stopping for ice-cream, it was his communication/social skills that were lacking. 

After r... (read more)

I also find the ice-cream example confusing because it's your main example but it doesn't seem like it supports your main points. For example replace Ash with a drug addict and ice cream with meth and Bryce suddenly looks like a hero trying to help his friend from making a big mistake. 

I think the below example still makes Bryce look like how it seems you intended him to look even if it was talking about meth. 

Ash: “Ooh, I want ice cream.”
Bryce: “Seriously?”
Ash: “Well, I'm not actually going to buy some, I just want some”
Bryce: “That’s the proble... (read more)

For anyone under 30, 10 years of polical lying, or 20 years for journalists "lying" is a long time and could be seen as a new reality, especially seeing as it seems to be working out quite well for the liars so it's unlikely to change anytime soon. 

Can you explain more, could a lower income worker without family nearby afford child care and full time house help?